mccann@cgdisis.cgd.ucar.edu (03/19/91)
Hi;
We are at the end of our ropes. Here's the situation: Our AST premium
286 computer running ANW 2.10 SFT II crashes about once a day with no warnings
or error messages. It is attached to an Ethernet network (via an NE1000)
that also has DECnet and TCP/IP running on it. We have 2 printers: an HPLJ
attached to the parallel port and an Apple LaserWriter with XON/XOFF on COM1.
We have 150MB and 650MB Storage Dimension disk drives attached to a Future
Domain SCSI controller.
The system has been running for about 3 years with an average of 15
users running wordprocessing and spreadsheet software, and making moderate
use of the printers. The frequency of the crashes seems to have increased
with the growth of our TCP/IP hosts that are attached to the network. We
have spent months looking for a "common denominator" to the crashes but
have found none. The crashes take 2 forms: the server will just re-boot
itself, or (more commonly) it will hang -- we can issue commands at
the console in this state (utilization usually stays fixed at ~5%, but has
on occasion gone up to 99%) but no one at a workstation can log in. Those
who were logged in get a "Network error reading drive f:" error and must
reboot. We (temorarily) fix it by just rebooting the fileserver.
We have swapped the CPU, network adapter, put in a power conditioner,
and moved the server to a different location, but the crashes continue.
Any advice you have will be greatly appreciated. I am particularly
interested in any evidence of an incompatibility with other protocols
on the ethernet or a problem with the NE1000 cards we have tried.
Thanks.
--
Mike McCann : Oceanography Department, Code OC/Mc - Naval Postgraduate School
Monterey, CA 93943 - (408) 646-3255 - mccann@whitney.oc.nps.navy.mil
"There's the one-armer now!" _ "new shoes"
- Gordon Cole / \/\ - Leo Johnson